17.2.1.2 Usage of Row-Based Logging and Replication

Row-based logging of temporary tables.  As noted in Section 17.4.1.30, “Replication and Temporary Tables”, temporary tables are not replicated when using row-based format or (from MySQL 8.0.4) mixed format. For more information, see Section 17.2.1.1, “Advantages and Disadvantages of Statement-Based and Row-Based Replication”.

Temporary tables are not replicated when using row-based or mixed format because there is no need. In addition, because temporary tables can be read only from the thread which created them, there is seldom if ever any benefit obtained from replicating them, even when using statement-based format.

You can switch from statement-based to row-based binary logging format at runtime even when temporary tables have been created. However, in MySQL 8.0, you cannot switch from row-based or mixed format for binary logging to statement-based format at runtime, because any CREATE TEMPORARY TABLE statements will have been omitted from the binary log in the previous mode.

The MySQL server tracks the logging mode that was in effect when each temporary table was created. When a given client session ends, the server logs a DROP TEMPORARY TABLE IF EXISTS statement for each temporary table that still exists and was created when statement-based binary logging was in use. If row-based or mixed format binary logging was in use when the table was created, the DROP TEMPORARY TABLE IF EXISTS statement is not logged. In releases before MySQL 8.0.4 and 5.7.25, the DROP TEMPORARY TABLE IF EXISTS statement was logged regardless of the logging mode that was in effect.

Nontransactional DML statements involving temporary tables are allowed when using binlog_format=ROW, as long as any nontransactional tables affected by the statements are temporary tables (Bug #14272672).

RBL and synchronization of nontransactional tables.  When many rows are affected, the set of changes is split into several events; when the statement commits, all of these events are written to the binary log. When executing on the slave, a table lock is taken on all tables involved, and then the rows are applied in batch mode. Depending on the engine used for the slave's copy of the table, this may or may not be effective.

Latency and binary log size.  RBL writes changes for each row to the binary log and so its size can increase quite rapidly. This can significantly increase the time required to make changes on the slave that match those on the master. You should be aware of the potential for this delay in your applications.

Reading the binary log.  mysqlbinlog displays row-based events in the binary log using the BINLOG statement (see Section 13.7.7.1, “BINLOG Syntax”). This statement displays an event as a base 64-encoded string, the meaning of which is not evident. When invoked with the --base64-output=DECODE-ROWS and --verbose options, mysqlbinlog formats the contents of the binary log to be human readable. When binary log events were written in row-based format and you want to read or recover from a replication or database failure you can use this command to read contents of the binary log. For more information, see Section 4.6.8.2, “mysqlbinlog Row Event Display”.

Binary log execution errors and slave_exec_mode.  Using slave_exec_mode=IDEMPOTENT is generally only useful with MySQL NDB Cluster replication, for which IDEMPOTENT is the default value. (See Section 22.6.10, “NDB Cluster Replication: Multi-Master and Circular Replication”). When slave_exec_mode is IDEMPOTENT, a failure to apply changes from RBL because the original row cannot be found does not trigger an error or cause replication to fail. This means that it is possible that updates are not applied on the slave, so that the master and slave are no longer synchronized. Latency issues and use of nontransactional tables with RBR when slave_exec_mode is IDEMPOTENT can cause the master and slave to diverge even further. For more information about slave_exec_mode, see Section 5.1.8, “Server System Variables”.

For other scenarios, setting slave_exec_mode to STRICT is normally sufficient; this is the default value for storage engines other than NDB.

Filtering based on server ID not supported.  You can filter based on server ID by using the IGNORE_SERVER_IDS option for the CHANGE MASTER TO statement. This option works with statement-based and row-based logging formats, but is deprecated for use when GTID_MODE=ON is set. Another method to filter out changes on some slaves is to use a WHERE clause that includes the relation @@server_id <> id_value clause with UPDATE and DELETE statements. For example, WHERE @@server_id <> 1. However, this does not work correctly with row-based logging. To use the server_id system variable for statement filtering, use statement-based logging.

Database-level replication options.  The effects of the --replicate-do-db, --replicate-ignore-db, and --replicate-rewrite-db options differ considerably depending on whether row-based or statement-based logging is used. Therefore, it is recommended to avoid database-level options and instead use table-level options such as --replicate-do-table and --replicate-ignore-table. For more information about these options and the impact replication format has on how they operate, see Section 17.1.6, “Replication and Binary Logging Options and Variables”.

RBL, nontransactional tables, and stopped slaves.  When using row-based logging, if the slave server is stopped while a slave thread is updating a nontransactional table, the slave database can reach an inconsistent state. For this reason, it is recommended that you use a transactional storage engine such as InnoDB for all tables replicated using the row-based format. Use of STOP SLAVE or STOP SLAVE SQL_THREAD prior to shutting down the slave MySQL server helps prevent issues from occurring, and is always recommended regardless of the logging format or storage engine you use.